Paradox about D's popularity.

[Jonathan M Davis]

On Wednesday, September 21, 2011 19:23:32 Gor F. Gyolchanyan wrote:
> As much as I'd like D to develop sweet language features like named function
> parameters, directing effort to increase compatibility of D with other
> languages (most notably, C++) would be a better idea, because if D gets
> cheaper and cheaper to use with existing code base, more and more people
> will start caring about D and the aforementioned sweet language features
> will get implemented more rapidly.
> 
> If Walter, Andrei and others agree with my proposal, i suggest to postpone
> all enhancement requests for D, except compatibility-related ones and
> direct all our attention to it. As well as direct more attention towards
> fixing DMD bugs, rather then developing phobos libraries, because as much
> as developing phobos is important, having a stable compiler is much more
> important.

With the release of TDPL, major feature changes to the language slowed down 
considerably. Features have been being added or changed primarily when it was 
deemed necessary to make features useable and them work as their supposed to 
(e.g. the addition of purity/@safe/nothrow inference for templates). Some 
smaller features have been added, but overall, the focus has been on fixing 
bugs in dmd and completing the implementation of existing features. So, 
nothing would really be gained by postponing enhancement requests at this 
point. Pretty much anything major has been put off until D3 already. Maybe some 
backwards-compatible features will be added to D2 later, but that sort of 
thing isn't generally being done right now. Rather, we've been getting a lot 
of bug fixes for dmd (and with the shift to github, the number and speed of bug 
fixes has increased considerably; the number of bug fixes for the last couple of 
releases has been huge).

- Jonathan M Davis